I’ve said it before and I’ll say it again! Development charges (“DCs”) are price signals, as such they influence decisions around land use and shape development patterns.

But DCs rarely follow the basic economic principle that, in order to achieve the proverbial “efficient allocation of resources” (read: efficient urban development patterns and infrastructure) prices should reflect costs.

In Ontario, DCs are inaccurate, because they don’t reflect the fact that urban form factors – type of development, density, location – are significant drivers of the cost of infrastructure. This is especially true for network or areal infrastructure, like roads, water and wastewater systems, stormwater management, and transit.

Instead, the typical approach is as follows. The infrastructure needed to support future population and employment growth is identified and costed. Then those costs are allocated to anticipated growth. For residential development, a cost per person is established. Then, this is multiplied by typical household sizes, to arrive a development charge per unit that varies by type.

The chart below shows an actual, typical charge.


The variations in DCs between unit types reflect the variations in average household sizes by type. Large apartments have a lower DC than detached singles because – and only because – they have an average household size of 2.5 people, compared with 3.5 people for the detached single.

But… on a per unit basis, the need for additional linear infrastructure created by an apartment is significantly lower than that caused by even a narrow-lot single detached house. This is for the obvious reason that as more units are added vertically to a given building footprint, land-related infrastructure costs per unit will fall.

Apartments represent a quantum shift in building form relative to houses, with many units sharing the same linear infrastructure. There is also a quantum (downward) shift in infrastructure costs per unit – a shift not reflected in the typical DC.

How might DC rates change if they were density-sensitive?
I undertook an exercise to explore how DCs rates might change if they reflected infrastructure cost variations related to different densities.

I took an actual, typical DC (the one shown above in the chart, from an upper-tier municipality in the Greater Golden Horseshoe). I used the same total infrastructure costs, unit counts and breakdowns by type that were used in determining the existing DC. But I reallocated the costs to include a factor that reflected the differing amounts of land required per single detached unit versus an apartment. In my view this measure, which reflects density, is a more accurate representation of cost causation for linear infrastructure than persons per unit.

So instead of the variation in DC levels between types being attributable to differing household sizes, my variations are attributable to differences in land intensity. In this way, the DC would reflect the efficiencies and lower per unit costs associated with the apartment form.

The density assumptions take the most conservative stance. Narrow lots were assumed for the detached singles and large apartment units in a low-rise, 3 to 4 storey building were assumed for the large apartments. Small apartments were assumed to be half the size of the large.

Population-related infrastructure costs by type were not changed, and are shown separately on the chart.


Implications of the exploration

1. When density is considered, DCs for apartments fall
The first observation is of course that the DCs for apartments go way down compared to the typical DC; they are reduced by about half , from $25,000 per unit to $12,500 per unit for a large apartment.

2. Conventional DCs overcharge apartments by significant amounts
This exercise suggests that apartments are overcharged, under this estimate, by 100%. Again, this is a conservative scenario, assuming the least dense apartments and most dense single lots. Apartment buildings with more units per hectare would be even more overcharged.

3. Overcharging of apartments under conventional DCs is the norm in Ontario
Given that almost all municipalities in Ontario use the household size method of cost allocation and ignore the effects of density on network infrastructure costs, the exercise suggests that significant overcharging of apartments is the norm.

We can expect that almost all new apartments in Ontario are overcharged by DCs. Some likely exceptions include: the City of Toronto, where almost all new growth consists of apartments, so there is little scope for the misallocation of costs between housing types; and the City of Markham, which uses a per hectare charge, which is inherently sensitive to density.

4. DCs for singles go up, by not by as much as one might expect
DCs for the single detached unit rise from about $35,000 to $40,000 per unit – an increase of $5,000. The reason that they don’t go up as much as the apartment DCs go down is that costs are reallocated from a smaller number of apartment units to a larger number of detached units, so the reallocated costs are spread over a larger number of detached units.

5. With a conventional DC, the apartment/single detached price differential is artificially reduced, suppressing demand for apartments
Under the conventional charge, the spread between DCs for apartments and DCs for single detached units is about $10,000.

Under the density-sensitive DC, the differential between the large apartment and the single detached unit increases to $28,000.


In other words, the conventional approach to DCs artificially reduces the cost differential between house types.

DCs are typically passed through to buyers in the final price of the home.

We often hear that the market price differential between apartments and houses is not significant enough to entice people into apartments. Conventional DCs are part of the reason why this is the case. They artificially make the prices of apartments and detached houses more similar than they would be if accurate DCs were used.

This makes it more difficult to market and sell apartments, which limits developers’ willingness to offer apartments on the market.

If reflected in house prices, the significant cost differential between house types inherent in a density-sensitive DC would increase the price differential between apartments and single detached units, making apartments a more attractive option.

6. At odds with Places to Grow
There are obvious implications of this financial mis-signaling for the Province’s Places to Grow plan, which seeks to densify and intensify. The financial misincentives created by inaccurate DCs discourage the kinds of development implied by Places to Grow and undermine the Plan’s objectives and policies.

7. Conventional DCs harbour inherent perverse subsidies
By ignoring the effects of density on infrastructure costs, the typical development charge in Ontario is structured in such as way as to create an inherent market distortion, a perverse subsidy in which apartments subsidize detached houses.

While there is nothing wrong with subsidies per se, good public policy requires that they be deliberate, transparent, and achieve public benefits.

The analysis shown here is only for one municipality; impacts on DC rates of taking density into account will vary from municipality to municipality, depending on a number of factors. This exercise gives an idea the potential implications.